| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Biomedical Sciences, University "G. D’Annunzio," Ce.S.I., 66013 Chieti, Italy; and
DNA Unit, Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Hospital Clínic, Universitat de Barcelona, Barcelona 08036, Spain
1 Correspondence: Laboratorio di Medicina Molecolare, Dipartimento di Scienze Biomediche, Università "G. D’Annunzio," Ce.S.I., Via dei Vestini, 31, 66013 Chieti, Italy. E-mail: mromano{at}unich.it
 |
ABSTRACT |
|---|
 TOP ABSTRACT THE CYCLOOXYGENASESCOX-2 AND CANCERTHE 5-LIPOXYGENASE5-LO AND CANCERCOX-2 AND 5-LO CONVERGING...CONCLUDING REMARKSREFERENCES |
|---|
Key Words: arachidonic acid • cancer • cyclooxygenase and lipoxygenase inhibitors • eicosanoids
|
THE CYCLOOXYGENASES |
|---|
|
TOPABSTRACTTHE CYCLOOXYGENASESCOX-2 AND CANCERTHE 5-LIPOXYGENASE5-LO AND CANCERCOX-2 AND 5-LO CONVERGING...CONCLUDING REMARKSREFERENCES |
|---|
, 2)
. A splice variant of COX-1, designated COX-3, that is especially sensitive to acetaminophen and related compounds has recently been identified (3)
. COX-1 and COX-2 are the products of two distinct genes, which in humans are localized on chromosomes 9 and 1, respectively (4
, 5)
. The COX-1 gene is 
22 kb in length with 11 exons and is transcribed as a 2.8 kb mRNA (4
, 6)
. The COX-2 gene is 8 kb long with 10 exons and is transcribed as 4.6, 4.0, and 2.8 kb mRNAs variants (2
, 7)
. Sequence analysis of the COX-2 5'-flanking region has revealed several potential transcription regulatory elements including a TATA box, a NF-IL-6 motif, two AP-2 sites, three Sp1 sites, two NF-
B sites, a CRE motif, and an E-box (8)
.
The amino acid sequences of COX-1 and COX-2 from a single species are 60% identical. COX-1 is a glycoprotein that in its native, processed form has 576 amino acids with an apparent molecular mass of 70 kDa (4
, 6)
. The cDNA for COX-2 encodes a polypeptide that before cleavage of the signal sequence contains 604 amino acids with an apparent molecular mass of 70 kDa (2
, 7)
.
COX-1 and COX-2 expression and tissue distribution are differentially regulated. COX-1 is ubiquitous and constitutively expressed throughout the gastrointestinal system, kidneys, vascular smooth muscle, and platelets (9)
. Conversely, COX-2 is undetectable in most tissues, but its expression can rapidly be induced by a variety of stimuli—growth factors and cytokines related to the inflammatory response (9)
. However, this distinction is not entirely accurate, since COX-1 can be induced or up-regulated under certain conditions and COX-2 has consistently been shown to be constitutively expressed in brain, kidney, as well as in megakaryocytes and newly formed platelets (9
, 10)
.
Immunoelectron microscopy and Western blot of COX-1 and COX-2 in subcellular fractions have revealed that COX-1 and COX-2 are located equally on the lumenal surfaces of the endoplasmic reticulum and nuclear envelope membranes (11)
.
Catalytic activity
COX transforms arachidonic acid into PGH2, the immediate precursor of various prostanoids, including prostaglandins (PGs) and thromboxane (TX). The COX pathway is of particular clinical relevance because it is the main target for drugs [such as nonsteroidal anti-inflammatory drugs (NSAIDs), COX-2 inhibitors (the so-called Coxibs) (12)
, and aspirin] used to relieve inflammation, pain, and fever and to prevent atherothrombosis. The basic catalytic activities of COX isozymes are similar enough to treat them as if they were biochemically identical. Accordingly, the biochemistry of these isozymes is discussed as one.
COX is a bifunctional enzyme that catalyzes the first two committed steps in the pathway leading to the formation of PGs and TX, namely, cyclooxygenation and peroxidation. In fact, COX cyclizes and adds two molecules of O2 to arachidonic acid to form the cyclic hydroperoxide PGG2. It also reduces PGG2 to PGH2. PGH2 is a highly unstable endoperoxide that functions as an intermediate substrate for the biosynthesis, by specific synthases and isomerases, of PGs of the E2, F2 and D2 series and also PGI2 (prostacyclin) and TXA2. These reactions are cell and tissue specific. Although COX-1 and COX-2 oxygenate arachidonate with almost identical kinetics, in general COX-2 is much more efficient with alternative substrates. For instance, COX-2 uses both fatty acids and endocannabinoids, including 2-archidonylglycerol (2-AG) and anandamide (AEA), as substrates. COX-2 action on 2-AG and AEA generates endocannabinoid-derived prostanoids, among them PG glycerol esters and ethanolamides (13)
. Recent studies indicate that aspirin-acetylated COX-2 converts arachidonic acid to 15R-hydroxyeicosatetraenoic acid (HETE), which is further transformed into 15-epi-lipoxins (14)
. Moreover, when exposed to aspirin, COX-2-expressing cells enzymatically transform omega-3 docosaexaenoic acid (DHA) to previously unrecognized compounds, i.e., a novel 17R series of hydroxy-DHAs (15)
. Both 15-epi-lipoxins and hydroxy-DHAs bear potent anti-inflammatory properties and play a key role in the resolution of inflammation. A schematic representation of these pathways is shown in Fig. 1
.
|
Prostanoids: receptors and functions
Prostanoids activate specific receptors, which are also cell and tissue specific. There are at least nine known PG receptors as well as several splice variants of them that belong to a subfamily of the G-protein-coupled receptor (GPCR) superfamily of seven transmembrane spanning proteins (16)
. Four of the receptor subtypes bind PGE2 (EP1-EP4), two bind PGD2 (DP1 and DP2) and the rest are single receptors for PGF2
, PGI2, and TXA2 (FP, IP, and TP, respectively) (16)
. IP, DP1, EP2, and EP4 receptors are coupled with the activation of Gs proteins and linked to increases in intracellular cAMP, whereas EP1, FP, and TP signal through Gq-mediated increases in intracellular calcium concentration. Exceptionally, EP3 receptor is coupled to Gi and decreases cAMP formation. PGs generated from COX-2 located in the nuclear membrane, in addition to signaling through G-protein-linked receptors, may control nuclear pathways through interactions with peroxisome proliferator-activating receptors (PPARs) (17)
. This emphasizes the importance of COX-2 as a regulator of nuclear events in cell growth and survival.
Activation of prostanoid receptors triggers a broad array of biological effects from the regulation of vascular tone, permeability, and platelet function to the induction of hyperalgesia and fever. See ref 18
for a detailed description of prostanoid bioactions.
|
COX-2 AND CANCER |
|---|
|
TOPABSTRACTTHE CYCLOOXYGENASESCOX-2 AND CANCERTHE 5-LIPOXYGENASE5-LO AND CANCERCOX-2 AND 5-LO CONVERGING...CONCLUDING REMARKSREFERENCES |
|---|
, COX-2 appears to be the isoform mainly involved in the pathogenesis of malignancy, particularly of colon cancer. The following are the most relevant findings obtained from several independent lines of research supporting the involvement of PGs and COX-2 in carcinogenesis.
Increased levels of PGs in tumors
The concentration of PGs, in particular PGE2, is increased in human colorectal cancer tissue when compared with normal colonic mucosa (20)
. Similar findings were reported in experimental carcinogenesis and in macroscopically normal colon of rodents exposed to colonic carcinogens (21)
.
Increased COX-2 expression in tumors
Consistent with elevated PGE2 levels, increased expression of COX-2, but not of COX-1, was detected in colorectal carcinomas (22
, 23)
. Published reports indicate that 85% of adenocarcinomas exhibit a 2- to 50-fold increase in COX-2 expression at both the mRNA and protein level compared with matched macroscopically normal colonic mucosal, being the rectum the site of maximal expression (22
, 23)
. However, the cellular origin of COX-2 overexpression in tumors is debated. Cancer cells, nontransformed epithelial cells, stromal fibroblasts, vascular endothelial cells, and inflammatory infiltrates are constituents of neoplastic colon tissue, and all of these cells are reported to display elevated COX-2 levels compared with the normal colon tissues (24)
. Other studies have shown that increased COX-2 expression in colon cancer originates mainly from the interstitial cells (i.e., macrophages), whereas little COX-2 expression was found in epithelial cancer cells (25)
.
Several mechanisms for enhanced COX-2 gene expression in cancer have been suggested, including mutations of APC and ras, activation of EGF receptor and IGF-I receptor pathways and heregulin/HER-2 receptor pathway, direct induction by the Epstein-Barr virus oncoprotein, and latent membrane protein 1 (26
, 27)
.
COX-2 overexpression promotes tumorigenesis; COX-2 inhibition is anti-neoplastic
The functional consequences of COX-2 overexpression have been analyzed mostly in the rat intestinal cell line (RIE cells), in the gastrocolonic cell lines HT-29 and HCA-7, and in the human colon cell line Caco-2 transfected with COX-2. In general, these experiments reveal that cells overexpressing COX-2 undergo phenotypic changes, including exhibition of increased adhesion to extracellular matrix proteins and resistance to apoptosis, which could enhance their tumorigenic potential (28)
. The proliferative activity of COX-2 is believed to be primarily mediated by PGs. In fact, the human gastrocolonic cancer cell line HT-29 shows increased proliferation in the presence of PGs in the culture medium (29)
. Moreover, in vivo, colonocyte proliferation is significantly stimulated by the injection of the stable derivative of PGE2, dimethyl PGE2, into normal mice (29)
. Consistently, transgenic COX-2 overexpression sensitizes mouse skin for tumor induction by carcinogens (30)
, whereas in APC
716 knockout mice, a model in which a targeted truncation deletion in the tumor suppressor gene APC causes intestinal polyposis, the number and size of the polyps were reduced dramatically (6- to 8-fold) when the COX-2 gene was knocked out (31)
.
Further evidence of the role played by COX-2 in cancer growth was provided by the use of NSAIDs and Coxibs. Sheng et al. tested the effects of a highly selective COX-2 inhibitor (SC-58125) on two transformed human colon cancer cell lines, only one of which had high levels of COX-2 protein expression and activity (32)
. These authors observed that SC-58125 decreased cell growth only in the COX-2-expressing cell line (32)
. In contrast, Elder et al. reported a significant anti-proliferative effect of NS-38, a selective COX-2 inhibitor, in colon cancer cells that do not express COX-2 (33)
. These controversial findings together with the observation that NSAIDs such as sulindac sulfone, a sulindac metabolite devoid of COX inhibitory activity, are able to reduce colon cancer cell growth (34)
suggest that COX-independent pathways could also be involved in the anti-proliferative effects of NSAIDs. A combination of epidemiological and experimental studies has demonstrated a clear association in vivo between regular long-term consumption of NSAIDs—aspirin in particular—and a reduced incidence in colon cancer (reviewed in ref 35
). Aspirin and sulindac have been also shown to reduce the number and size of adenomatous colonic polyps in patients with familial adenomatous polyposis (FAP) (36)
. Epidemiological data in humans are well documented, because separate studies differing in location, design, and population have consistently shown that regular use of NSAIDs reduces the risk of colon cancer. Parallel studies in animal models of colon carcinogenesis have also proved that aspirin as well as other conventional NSAIDs such as piroxicam, indomethacin, sulindac, ibuprofen, and ketoprofen inhibits chemically induced colon cancer in rats and mice (reviewed in ref 35
). In these studies, the mechanism by which NSAIDs reduce the risk of cancer is likely to be related to the inhibition of COX-2. A randomized clinical trial has shown that celecoxib significantly inhibits the growth of adenomatous polyps and causes regression of existing polyps in patients with hereditary FAP (37)
. Studies in rodents demonstrated that selective pharmacological inhibition of COX-2 activity prevents chemically induced carcinogenesis and intestinal polyp formation in an experimental model of FAP (38
, 39)
. In Min mice and rats exposed to chemical carcinogens, selective COX-2 inhibitors induced apoptosis and suppressed growth of a variety of epithelial cancers (reviewed in ref 40
). COX-2 may also be implicated in head and neck, lung, and pancreatic tumors, and the beneficial effects of Coxibs may be extended to these cancers (40
41
42)
. Along these lines, COX-2 inhibition appears to potentiate antitumor activity of chemotherapeutic agents (43)
. This could be related to the recently discovered capability of COX-2 to block p53- or genotoxic stress-induced apoptosis (44)
.
COX-2 induces neo-angiogenesis
Recent results indicate that neo-angiogenesis, which is essential for tumor development, requires COX-2 (45)
. In fact, COX-2 overexpressing colon cancer cells produce large amounts of proangiogenic factors, including vascular endothelial growth factor (VEGF) (46)
, a key regulator of endothelial cell migration and in vitro angiogenesis. Also, tumors implanted in COX-2 (-/-) mice display a reduction in vascular density and growth (47)
. This may be in relation with the inability of COX-2 (-/-) stromal fibroblasts to produce VEGF (47)
. A relationship between COX-2 and VEGF expression can be also observed in prostate cancer cells and human breast cancer (48
, 49)
. Furthermore, PGs induce expression of VEGF and of the VEGF transcription factor hypoxia-inducible factor-1 (50
, 51)
. Nevertheless, a possible role for COX-1 in the regulation of angiogenesis cannot be excluded, since treatment of endothelial cells with either aspirin or COX-1 anti-sense oligonucleotides inhibits endothelial tube formation (46)
. Using the dual COX-1/-2 inhibitor diclofenac, Seed et al. showed that colorectal cancer growth is inhibited via the suppression of angiogenesis (52)
. On the other hand, more recent studies demonstrated that selective COX-2, but not COX-1, inhibitors suppress the growth of corneal capillary blood vessels in rats exposed to basic fibroblast growth factor and inhibit the growth of several human tumors implanted in mice (47
, 53)
.
|
THE 5-LIPOXYGENASE |
|---|
|
TOPABSTRACTTHE CYCLOOXYGENASESCOX-2 AND CANCERTHE 5-LIPOXYGENASE5-LO AND CANCERCOX-2 AND 5-LO CONVERGING...CONCLUDING REMARKSREFERENCES |
|---|
, is a 78 kDa protein that catalyzes the biosynthesis of potent bioactive eicosanoids. The 5-LO gene is > 82 kb in length and encompasses 14 exons separated by 13 introns (55)
. The 5'-flanking region of this gene displays a unique (G+C)-rich sequence encompassing five tandem SP1 consensus regions, but not TATA or CCAAT sequences. The -179 to -56 DNA region appears to be essential for transcription initiation (56)
. The 5-LO promoter contains consensus regions for a number of transcription regulators belonging to the Egr, Sp, NF-B, GATA, myb, and AP families (56)
. Mutations characterized by alterations in the number of Sp1 and Egr-1 binding motifs associated with reduced reporter assay activity have been detected, although their pathophysiological role remains to be established (57)
.
In humans, 5-LO is physiologically expressed in cells of the myeloid lineage, but also in B lymphocytes and pulmonary artery endothelial cells. However, the mechanisms that selectively regulate the degree of 5-LO gene expression remain elusive. It is known that transforming growth factor ß and 1,25-dihydroxyvitamin D3 increase 5-LO mRNA accumulation and 5-LO activity (58)
, but a direct effect on 5-LO transcription by these agents has not been demonstrated. On the other hand, the 5-LO promoter is repressed by methylation of CpG sites within the (G+C)-rich sequence (59)
. It has been proposed that DNA demethylation may represent a key prerequisite for full 5-LO expression (59)
. Given the variety of putative transcription regulators of the 5-LO promoter, it is reasonable to assume that control of 5-LO expression may be quite complex and multifactorial.
Catalytic activity and intracellular localization
As other members of the LO family, 5-LO is a dioxygenase. It stereospecifically removes the pro-S hydrogen at carbon-7 of one of the 1,4-cis,cis-pentadiene structures present in its natural substrate, arachidonic acid. This reaction is accompanied by the insertion of molecular O2 at carbon-5, giving rise to 5(S)-hydroperoxy-6E,8Z,11Z,14Z-eicosatetraenoic acid (HPETE). 5(S)-HPETE can be reduced to 5(S)-hydroxy-6E,8Z,11Z,14Z-eicosatetraenoic acid (HETE), which can be in turn dehydrogenated to 5-oxo-HETE. 5(S)-HPETE can also be subjected to the stereospecific removal of the pro-R hydrogen at carbon-10 to generate the highly unstable allylic epoxide 5(S),6(S)-oxido-7E,9E,11Z,15Z-eicosatetraenoic acid or leukotriene (LT)A4. This can be further metabolized by multiple routes, the best-characterized of which follow. 1) Conversion to LTB4 by LTA4 hydrolase (60)
. 2) Transformation to LTC4 by LTC4 synthase (61)
. LTC4 is further metabolized to LTD4 and LTE4. Together, LTC4, D4, and E4 correspond to the slow-reacting substance of anaphylaxis (SRS-A) and are now termed cysteinyl leukotrienes (Cys-LTs). 3) Conversion to lipoxin (LX) A4 and B4 by 12- or 15-LO (62)
. A schematic representation of these pathways is shown in Fig. 2
.
|
For full activity 5-LO requires cofactors such as calcium, ATP, and a 18 kDa protein termed 5-LO-activating protein (FLAP), which facilitates the docking of arachidonic acid to 5-LO (63)
; post-translational modifications, i.e., phosphorylation by p38 kinase-dependent MAPKAP kinases (64)
; translocation to the nuclear envelope (65)
. This latter represents a unique feature of 5-LO not shared by other arachidonic acid catalyzing enzymes. In fact, 5-LO may be localized either in the cytoplasm or in the nuclear euchromatin, depending on the cell type, but it is always translocated to the nuclear membrane upon cell activation (65)
. It is at this site that 5-LO encounters phospholipase A2 and FLAP and where biosynthesis of 5-LO products is initiated (66)
. Nuclear import of 5-LO is crucial for 5(S)-HETE and LT biosynthesis, as 5-LO mutants lacking sequences involved in nuclear targeting are catalytically inactive (67)
. Recently, a basic region on human 5-LO encompassing residues 518-530 has been recognized to meet the criteria of a nuclear import sequence (NIS) (68)
.
5-LO products: receptors and functions
The most relevant pathophysiological role recognized to 5-LO products is regulation of the immune inflammatory response. LTB4 is in fact a potent activator of neutrophil chemotaxis and trans-endothelial migration, whereas the Cys-LTs, are key mediators of allergic inflammation. 5(S)-HETE and 5-oxo-HETE also activate neutrophils and/or monocytes. Specific receptors for LTB4 and Cys-LT have recently been cloned (69
70
71
72)
. More recently, a Gi/o-coupled eicosanoid receptor, recognized by 5-oxo-HETE has been identified (73)
. These findings represent a significant advance for a better understanding of the relevance of individual 5-LO products in health and disease. We already know that 5-LO knockout mice live normally and do not show appreciable dysfunctions; they are resistant to selected inflammatory damage, but more sensitive to bacteremia (74
, 75)
. Genetic manipulation of receptors for 5-LO products may provide new insight into the biological significance of individual 5-LO pathways.
|
5-LO AND CANCER |
|---|
|
TOPABSTRACTTHE CYCLOOXYGENASESCOX-2 AND CANCERTHE 5-LIPOXYGENASE5-LO AND CANCERCOX-2 AND 5-LO CONVERGING...CONCLUDING REMARKSREFERENCES |
|---|
5-LO is overexpressed in cancer cells and human tumors
In addition to normal cells, 5-LO is expressed by a broad variety of cancer cells including colon, lung, breast, prostate, pancreas, bone, brain, and mesothelium (76
77
78
79
80
81
82)
. Human cancer tissues also display an enhanced expression of 5-LO (83)
. At least for human mesothelioma and pancreatic cancer, very low or undetectable 5-LO expression is observed in normal cells or tissues, whereas high 5-LO expression and activity are detected in the corresponding transformed cells or tissues (81
, 82)
. Thus, 5-LO expression appears to be occasionally up-regulated during neoplastic transformation, although the sequelae of events connecting cancer development and 5-LO gene expression are unknown. Whether mutations of the 5-LO promoter region occur in cancer and whether oncogenes induce transcription of 5-LO remain to be investigated. Notably, expression of receptors for 5-LO products may also be relevant in cancer development. Indeed, CysLT1 is highly expressed in colorectal adenocarcinomas and its degree of expression correlates negatively with patient survival (84)
.
5-LO promotes cancer cell growth and neo-angiogenesis
The biological functions of 5-LO in cancer cells have been examined using pharmacological inhibitors and/or antisense technology. Since the mid-1980s, when the anti-proliferative effects of 5-LO inhibitors on cancer cell lines began to be reported (85)
, a considerable number of studies have confirmed that 5-LO activity promotes cancer cell proliferation and survival. How this is achieved is not fully understood, although regulation of growth factor expression may be involved. We reported that 5-LO products, namely, 5(S)-HETE and LTA4 but not LTB4, potently up-regulate VEGF transcription in a human malignant mesothelioma model (82)
. VEGF is a potent autocrine growth factor for mesothelioma cells and its selective suppression, achieved with 5-LO antisense oligonucleotides and/or 5-LO inhibitors, results in apoptotic cell death (82)
. Since VEGF is also a potent proangiogenic factor and neo-angiogenesis and is crucial for tumor growth and invasion, it might be concluded that 5-LO promotes in vivo tumor development by a dual mechanism, a direct proliferative stimulus on cancer cells and a potentiation of the proangiogenic response by the host stromal cells. Indeed, 5-LO products up-regulate VEGF transcription in human umbilical vein endothelial cells (M. Romano et al., unpublished observation). On the other hand, 5-LO activity appears to interfere with the mechanisms of tumor suppression (A. Catalano et al., unpublished observation), suggesting that this enzyme may condition tumor response to genotoxic therapeutic agents.
|
COX-2 AND 5-LO CONVERGING PATHWAYS IN CANCER |
|---|
|
TOPABSTRACTTHE CYCLOOXYGENASESCOX-2 AND CANCERTHE 5-LIPOXYGENASE5-LO AND CANCERCOX-2 AND 5-LO CONVERGING...CONCLUDING REMARKSREFERENCES |
|---|
). First, COX-2 and 5-LO are coexpressed and up-regulated in a quite large number of cancer cells and human tumors, including lung, colon, prostate, breast, and mesothelioma (76
, 82)
. Second, both 5-LO and COX-2 are proangiogenic with a convergent targeting on VEGF expression and release (46
47
48
49
50
51
, 82)
. Third, COX-2, as well as 5-LO inhibitors, arrest cell cycle progression and induce apoptotic cell death in a number of cancer cells (77
, 78
, 82
, 85
, 86)
. Fourth, both active COX-2 and 5-LO are localized in the nucleus (11
, 65)
; COX-2 and 5-LO products may function as endogenous ligands for nuclear receptors as the PPARs, although this issue raises some controversy (87)
. Fifth, both COX-2 and 5-LO are up-regulated by genotoxic stress and inhibit p53-induced cell death (ref 44
and A. Catalano et al., unpublished observation).
|
Functional interactions and therapeutic implications
COX-2 and 5-LO may have redundant functions in cancer pathobiology. Can we hypothesize that when coexpressed these arachidonic acid-metabolizing enzymes represent an integrated system that regulate the proliferative, proangiogenic potential of cancer cells? Should we include in this system other LOs (i.e., the 12 or 15 isoforms) that are often expressed in cancer cells? (76)
. Finally, how does this system work? Information on arachidonic acid metabolism in cancer is based largely on pharmacological studies. As outlined earlier in this review, COX-2 selective and nonselective inhibitors have been and are being used both in vitro and in vivo to block cancer progression. Animal studies and clinical trials with these drugs are encouraging, particularly in FAP and colorectal polyposis (35)
, although are still pending as to their side effects and precise mechanism(s) of action. NSAIDs may induce cancer cell apoptosis independent of COX-2 inhibition. For example, in esophageal cancer cells NSAIDs-induced apoptosis seems to depend on 15-LO up-regulation (88)
. On the other hand, a mechanism linking COX-2 inhibition to apoptosis could be an increase in concentration of free, unmetabolized arachidonic acid (89)
. According to this hypothesis, the blockade of arachidonic metabolism may increase the intracellular levels of unesterified arachidonate, which itself induces a concentration-dependent apoptotic response. If this is the case, when multiple metabolic pathways of arachidonic acid are present within the same cell, as in cancer cells coexpressing COX-2 and 5-LO (76)
, both enzymes should be blocked to achieve a substantial elevation in free arachidonic acid levels. Nevertheless, in these cells the simultaneous inhibition of COX-2 and 5-LO may prevent the shunting of arachidonate toward 5-LO when COX-2 is blocked by NSAIDs, thus suppressing the production of 5-LO-derived mitogenic proangiogenic eicosanoids such as 5-oxo-HETE, 5(S)-HETE, and LTA4. However, this may represent an oversimplistic representation of the possible interactions between LO and COX pathways in cancer cells. In fact, 12-LO and 15-LO are also expressed in cancer cells, although at a lower frequency than COX-2 and 5-LO. 12-LO has actions similar to 5-LO and COX-2, since it stimulates proliferation and may be proangiogenic (90)
, whereas 15-LO isoforms may have opposite effects on cancer cells, being the isoform 1 stimulatory and the isoform 2 inhibitory of cancer cell proliferation (91)
. Therefore, an ideal cancer treatment targeted to the regulation of arachidonic metabolism should block COX-2, 5-LO, 12-LO, and 15-LO-1 without altering or even inducing 15-LO-2. No drug that recapitulates these capabilities is currently available, although combined COX/5-LO inhibitors have been obtained. The best-characterized are E34122 (92)
, S-2474 (93)
, tryptanthrin (94)
, licofelone (ML-3000) (95)
, and a recently synthesized methoxytetrahydropyran derivative (96)
. These drugs have been tested in vitro and in animal models of inflammatory disease with encouraging results. Licofelone is now in phase III clinical trials. Compared with single inhibitors, this class of drug has the advantage of an improved anti-inflammatory action associated with a reduction in side effects (97)
. These molecules have not yet been used to treat cancer patients, although in vitro evidence of the role played by COX-2 and 5-LO in cancer progression would justify such a therapeutic approach. 5-LO inhibitors are also available for clinical use (98)
. They could be administered in association with NSAIDs to achieve a more extensive block of production of arachidonic acid metabolites that may promote cancer progression. It should be pointed out that COX-2 inhibition sensitizes cancer cells to routinely used chemotherapeutic agents (43)
. A similar effect can be observed with 5-LO inhibition (unpublished results). Thus, dual COX-2/5-LO inhibitors may also help improve the efficacy of conventional anticancer treatments and reduce their side effects.
This approach should take into account the possibility that genetic variants of COX-2 and 5-LO may be associated with cancer development and/or resistance to inhibitors. Our current knowledge of pharmacogenetics and pharmacogenomics of arachidonic acid metabolizing enzymes is limited, but research in this field is progressing. It has been reported that a COX-2 variant, Val511Ala, can be associated with a reduced risk of colorectal neoplasia (99)
. Individuals heterozygous for the A-842G/C50T haplotype of COX-1 are more sensitive to aspirin (100)
. It is predicted that in the near future selection of cancer patients for treatment with anti-COX-2/5-LO drugs will be also based on genetic analyses of these enzymes.
|
CONCLUDING REMARKS |
|---|
|
TOPABSTRACTTHE CYCLOOXYGENASESCOX-2 AND CANCERTHE 5-LIPOXYGENASE5-LO AND CANCERCOX-2 AND 5-LO CONVERGING...CONCLUDING REMARKSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication March 17, 2003. Accepted for publication June 11, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTTHE CYCLOOXYGENASESCOX-2 AND CANCERTHE 5-LIPOXYGENASE5-LO AND CANCERCOX-2 AND 5-LO CONVERGING...CONCLUDING REMARKSREFERENCES |
|---|
